{"id":2155,"date":"2021-07-14T09:30:00","date_gmt":"2021-07-14T07:30:00","guid":{"rendered":"http:\/\/cirpicme.org\/?page_id=2155"},"modified":"2021-07-13T18:57:28","modified_gmt":"2021-07-13T16:57:28","slug":"unsupervised-domain-adaptive-object-detection-for-assembly-quality-inspection","status":"publish","type":"page","link":"https:\/\/cirpicme.org\/index.php\/assembly-battery-production\/unsupervised-domain-adaptive-object-detection-for-assembly-quality-inspection\/","title":{"rendered":"Unsupervised domain adaptive object detection for assembly quality inspection"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\"><em>by <em>Xiaomeng Zhu, Atsuto Maki, Lars Hanson<\/em><\/em> <em>(Sweden)<\/em><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Abstract<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">A challenge to apply deep learning-based computer vision technologies for assembly quality inspection lies in the diverse assembly approaches and the restricted annotated training data. This paper describes a method for overcoming the challenge by training an unsupervised domain adaptive object detection model on annotated synthetic images generated from CAD models and unannotated images captured from cameras. On a case study of pedal car front-wheel assembly, the model achieves promising results compared to other state-of-the-art object detection methods. Besides, the method is efficient to implement in production as it does not require manually annotated data.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Keywords<\/strong>: Domain adaptation, Object detection, Automatic assembly quality inspection, Deep learning<\/p>\n\n\n\n<div style=\"height:20px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Video presentation<\/strong><\/p>\n\n\n\n<figure class=\"wp-block-video\"><video height=\"1080\" style=\"aspect-ratio: 1920 \/ 1080;\" width=\"1920\" controls src=\"http:\/\/cirpicme.org\/wp-content\/uploads\/2021\/07\/Xiaomeng_Zhu.mp4\"><\/video><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\"><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Presenting author<\/strong><\/p>\n\n\n\n<figure class=\"wp-block-table is-style-stripes\"><table class=\"has-subtle-light-gray-background-color has-background\"><tbody><tr><td><\/td><td><\/td><td><\/td><\/tr><tr><td><img data-recalc-dims=\"1\" loading=\"lazy\" decoding=\"async\" width=\"150\" height=\"150\" class=\"wp-image-1862\" style=\"width: 150px;\" src=\"https:\/\/i0.wp.com\/cirpicme.org\/wp-content\/uploads\/2021\/07\/Xiaomeng_Zhu_Photo.jpg?resize=150%2C150\" alt=\"\" srcset=\"https:\/\/i0.wp.com\/cirpicme.org\/wp-content\/uploads\/2021\/07\/Xiaomeng_Zhu_Photo.jpg?w=220&amp;ssl=1 220w, https:\/\/i0.wp.com\/cirpicme.org\/wp-content\/uploads\/2021\/07\/Xiaomeng_Zhu_Photo.jpg?resize=150%2C150&amp;ssl=1 150w\" sizes=\"auto, (max-width: 150px) 100vw, 150px\" \/><\/td><td><strong>Name:<\/strong><br><br><strong>Affiliation:<\/strong><br><br><strong>Email:<\/strong><\/td><td>Xiaomeng Zhu<br><br>Scania CV AB and KTH Royal Institute of Technology, Sweden<br><br>xiazhu@kth.se<\/td><\/tr><tr><td><\/td><td><\/td><td><\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\"><\/p>\n","protected":false},"excerpt":{"rendered":"<p>by Xiaomeng Zhu, Atsuto Maki, Lars Hanson (Sweden) Abstract A challenge to apply deep learning-based computer vision technologies for assembly quality inspection lies in the diverse assembly approaches and the restricted annotated training data. This paper describes a method for overcoming the challenge by training an unsupervised domain adaptive object&#8230;<\/p>\n<p> <a class=\"continue-reading-link\" href=\"https:\/\/cirpicme.org\/index.php\/assembly-battery-production\/unsupervised-domain-adaptive-object-detection-for-assembly-quality-inspection\/\"><span>Continue reading<\/span><i class=\"crycon-right-dir\"><\/i><\/a> <\/p>\n","protected":false},"author":9,"featured_media":0,"parent":2406,"menu_order":0,"comment_status":"open","ping_status":"closed","template":"","meta":{"nf_dc_page":"","om_disable_all_campaigns":false,"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"footnotes":""},"class_list":["post-2155","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/cirpicme.org\/index.php\/wp-json\/wp\/v2\/pages\/2155","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/cirpicme.org\/index.php\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/cirpicme.org\/index.php\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/cirpicme.org\/index.php\/wp-json\/wp\/v2\/users\/9"}],"replies":[{"embeddable":true,"href":"https:\/\/cirpicme.org\/index.php\/wp-json\/wp\/v2\/comments?post=2155"}],"version-history":[{"count":1,"href":"https:\/\/cirpicme.org\/index.php\/wp-json\/wp\/v2\/pages\/2155\/revisions"}],"predecessor-version":[{"id":2156,"href":"https:\/\/cirpicme.org\/index.php\/wp-json\/wp\/v2\/pages\/2155\/revisions\/2156"}],"up":[{"embeddable":true,"href":"https:\/\/cirpicme.org\/index.php\/wp-json\/wp\/v2\/pages\/2406"}],"wp:attachment":[{"href":"https:\/\/cirpicme.org\/index.php\/wp-json\/wp\/v2\/media?parent=2155"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}